[marandaneto]

💡 Motivation and Context

NaN and positive/negative infinity can make JSON encoding discard an event or an entire batch. Following the design checkpoint and explicit maintainer approval, this revision narrows the original proposal to bounded, data-only payloads instead of reproducing PHP object serialization.

• Repair non-finite numbers to JSON null only when the whole payload contains scalar values/arrays, no objects/resources/retained array-element references, valid strings, at most 10,000 values, 64 array levels, and 1 MiB of raw string/key bytes. Counts include the complete envelope and SDK metadata.
• Otherwise pass the original whole payload unchanged to native JSON. Exhausting an inspection budget is not a new rejection: finite payloads still encode natively. Objects, enums, serializer results and reference-cell updates are not normalized.
• This deliberately does not isolate events in a mixed batch. An opaque value anywhere can leave a non-finite neighbor subject to native encoding failure. Ineligible File events retain existing failure handling, including empty-line output. The encoded batch-size guard is unchanged and separate from the repair budget.
• Remove the custom sequential writer, object-handler emulation, property-layout reconstruction, active-reference tracking and synthetic JSON error restoration. Eligibility is read-only; copying starts only after full qualification. Native JSON remains authoritative for ordering, redaction, callback counts, recursion, exceptions and errors. There are no serializer retries, partial-output flags or token replacement.

Existing SDK methods/options are unchanged; capture callers can use the existing before_send hook for their own JSON-safe conversion policy. The encoder lives in PostHog\Internal and that namespace is excluded from public API snapshots. The API snapshot and README match main; no public API is added. The serialization/design docs were removed at maintainer request, with the agreed limits and fallback policy retained in this description. A patch change intent is included. The capture spec has no NaN/infinity policy. Related Python work: PostHog/posthog-python#1081.

💚 How did you test it?

• Retained every concrete reported reproduction. The five newest cases failed on 577a8b7 and now pass without changing their expectations: abstract-property redaction order, stateless/stateful active-array serializer re-entry, hooked DateInterval output/caller state, and a 12 MiB string under 32 MiB before normal batch-size rejection.
• The new data-only policy had six expected failures against the previous writer before replacement. Boundary tests now cover exact value/byte/depth limits, successful finite fallback beyond each bound, opaque/reference whole-payload fallback, caller immutability and no serializer retry.
• Original object/non-finite examples remain as native-failure cases under the approved narrowed contract. Added finite companions retain successful wire-format/privacy coverage rather than dropping the examples. Added supported data-only NaN/infinity integration cases for both batch and File consumers, plus 32 MiB dense-array and escaped-string repair probes against independent native-null fixtures.
• Follow-up retained-capacity repros were red on d16a8b1: a million-slot pruned root, 16 repeated pruned children, and a sparse high key each exhausted 32 MiB while native-error/native-null baselines completed normally. Key-preserving compaction now passes all three under the same limit; a numeric-string symbol-table regression also protects object-vs-list JSON shape. Unchanged prefixes, sparse keys and caller values remain covered.
• 165 focused tests / 1,220 assertions and the complete 1,595-test / 8,349-assertion suite with Xdebug 3.5.3 coverage pass locally on PHP 8.5.10. Existing notices remain 27 deprecations and 42 PHPUnit deprecations. All 16 adapter tests pass.
• An API snapshot regression was red before excluding the internal namespace and now passes, while retaining existing public/legacy symbols. All encoding regressions remain unchanged apart from imports and fully qualified helper references.
• CI runs PHPUnit on PHP 8.2, 8.3, 8.4 and 8.5. Changed-file PHP_CodeSniffer, public API snapshot checks, strict Composer validation and whitespace checks pass. Tests use offline fakes/loopback servers, not live PostHog endpoints.

Compliance assertions remain advisory: lib_curl 46/47, socket 45/47 and fork_curl 36/47 on d16a8b1, matching the previous 577a8b7 reports exactly. Inventory-completeness checks pass; unrelated transport gaps are not changed here.

Design invariants were reviewed separately from test counts: no copying before full eligibility; no executable user values/reference cells in a repairable graph; original graphs preserved for native fallback; primitive string validation does not overwrite callback-visible JSON error state. The initial assumption that live entry counts bound copy-on-write capacity was disproved; normalization now compacts live entries before mapping values and reuses unchanged subtrees without writing into caller-seeded tables. Earlier autoreview covered an older revision, not this replacement; human review remains required.

📝 Checklist

• I reviewed the submitted code.
• I added tests to verify the changes.
• I updated the docs if needed.
• No breaking change or entry added to the changelog.

If releasing new changes

• Ran pnpm change to generate a change intent file

🤖 Agent context

Autonomy: Human-driven (agent-assisted)

Human DRI: @marandaneto. Implemented with Pi using bash and file tools. Earlier graph/writer approaches were rejected after compatibility and ownership regressions; the maintainer approved conservative whole-payload native delegation with narrowly bounded data-only repair. No new hook or SDK option was introduced. No public session link is available. Human review is required.

[posthog]

PostHog Review · Standard · reviewed a3fd115: 1 Must fix, 1 Should fix.

This is the only automatic review while this pull request is a draft. Comment @posthog review to review new changes sooner. Otherwise, a review of the new changes runs when you mark it ready for review.

[github-actions]

posthog-php-socket Compliance Report

Date: 2026-10-23T13:00:30.216518+00:00
Duration: 114637ms

⚠️ Some Tests Failed

45/47 tests passed, 2 failed

---

Capture Tests

⚠️ 28/30 tests passed, 2 failed

View Details

Test	Status	Duration
Format Validation.Event Has Required Fields	✅	25ms
Format Validation.Event Has Uuid	✅	520ms
Format Validation.Event Has Lib Properties	✅	524ms
Format Validation.Distinct Id Is String	✅	520ms
Format Validation.Token Is Present	✅	522ms
Format Validation.Custom Properties Preserved	✅	520ms
Format Validation.Event Has Timestamp	✅	520ms
Format Validation.Non Utc Event Timestamp Is Converted To Utc	✅	522ms
Retry Behavior.Retries On 503	✅	5832ms
Retry Behavior.Does Not Retry On 400	✅	2524ms
Retry Behavior.Does Not Retry On 401	✅	2523ms
Retry Behavior.Respects Retry After Header	❌	5631ms
Retry Behavior.Implements Backoff	✅	16238ms
Retry Behavior.Retries On 500	✅	5137ms
Retry Behavior.Retries On 502	✅	5628ms
Retry Behavior.Retries On 504	✅	5632ms
Retry Behavior.Max Retries Respected	✅	16246ms
Deduplication.Generates Unique Uuids	✅	527ms
Deduplication.Preserves Uuid On Retry	✅	5629ms
Deduplication.Preserves Uuid And Timestamp On Retry	✅	10838ms
Deduplication.Preserves Uuid And Timestamp On Batch Retry	✅	5633ms
Deduplication.No Duplicate Events In Batch	✅	527ms
Deduplication.Different Events Have Different Uuids	✅	523ms
Compression.Sends Gzip When Enabled	✅	522ms
Batch Format.Uses Proper Batch Structure	✅	522ms
Batch Format.Flush With No Events Sends Nothing	✅	518ms
Batch Format.Multiple Events Batched Together	✅	511ms
Error Handling.Does Not Retry On 403	✅	2524ms
Error Handling.Does Not Retry On 413	✅	2524ms
Error Handling.Retries On 408	❌	5529ms

Failures

retry_behavior.respects_retry_after_header

Retry delay too short: 103ms < 2500ms

error_handling.retries_on_408

Expected at least 2 requests, got 1

Feature_Flags Tests

✅ 17/17 tests passed

View Details

Test	Status	Duration
Request Payload.Request With Person Properties Device Id	✅	520ms
Request Payload.Flags Request Uses V2 Query Param	✅	524ms
Request Payload.Flags Request Hits Flags Path Not Decide	✅	518ms
Request Payload.Flags Request Omits Authorization Header	✅	519ms
Request Payload.Token In Flags Body Matches Init	✅	519ms
Request Payload.Groups Round Trip	✅	524ms
Request Payload.Groups Default To Empty Object	✅	520ms
Request Payload.Disable Geoip False Propagates As Geoip Disable False	✅	520ms
Request Payload.Disable Geoip Omitted Defaults To False	✅	519ms
Request Payload.Flag Keys To Evaluate Contains Only Requested Key	✅	520ms
Request Lifecycle.No Flags Request On Init Alone	✅	516ms
Request Lifecycle.No Flags Request On Normal Capture	✅	508ms
Request Lifecycle.Two Flag Calls Produce Two Remote Requests	✅	527ms
Request Lifecycle.Mock Response Value Is Returned To Caller	✅	522ms
Retry Behavior.Retries Flags On 502	✅	621ms
Retry Behavior.Retries Flags On 504	✅	621ms
Side Effect Events.Get Feature Flag Captures Feature Flag Called Event	✅	523ms

[github-actions]

posthog-php-fork_curl Compliance Report

Date: 2026-10-23T13:00:19.534237+00:00
Duration: 112407ms

⚠️ Some Tests Failed

36/47 tests passed, 11 failed

---

Capture Tests

⚠️ 19/30 tests passed, 11 failed

View Details

Test	Status	Duration
Format Validation.Event Has Required Fields	✅	25ms
Format Validation.Event Has Uuid	✅	520ms
Format Validation.Event Has Lib Properties	✅	522ms
Format Validation.Distinct Id Is String	✅	522ms
Format Validation.Token Is Present	✅	521ms
Format Validation.Custom Properties Preserved	✅	522ms
Format Validation.Event Has Timestamp	✅	523ms
Format Validation.Non Utc Event Timestamp Is Converted To Utc	✅	524ms
Retry Behavior.Retries On 503	❌	5524ms
Retry Behavior.Does Not Retry On 400	✅	2528ms
Retry Behavior.Does Not Retry On 401	✅	2524ms
Retry Behavior.Respects Retry After Header	❌	5527ms
Retry Behavior.Implements Backoff	❌	15538ms
Retry Behavior.Retries On 500	❌	5526ms
Retry Behavior.Retries On 502	❌	5534ms
Retry Behavior.Retries On 504	❌	5526ms
Retry Behavior.Max Retries Respected	❌	15538ms
Deduplication.Generates Unique Uuids	✅	527ms
Deduplication.Preserves Uuid On Retry	❌	5527ms
Deduplication.Preserves Uuid And Timestamp On Retry	❌	10533ms
Deduplication.Preserves Uuid And Timestamp On Batch Retry	❌	5529ms
Deduplication.No Duplicate Events In Batch	✅	528ms
Deduplication.Different Events Have Different Uuids	✅	523ms
Compression.Sends Gzip When Enabled	✅	526ms
Batch Format.Uses Proper Batch Structure	✅	523ms
Batch Format.Flush With No Events Sends Nothing	✅	514ms
Batch Format.Multiple Events Batched Together	✅	516ms
Error Handling.Does Not Retry On 403	✅	2525ms
Error Handling.Does Not Retry On 413	✅	2525ms
Error Handling.Retries On 408	❌	5528ms

Failures

retry_behavior.retries_on_503

Expected at least 3 requests, got 1

retry_behavior.respects_retry_after_header

Expected at least 2 requests, got 1

retry_behavior.implements_backoff

Expected at least 3 requests, got 1

retry_behavior.retries_on_500

Expected at least 2 requests, got 1

retry_behavior.retries_on_502

Expected at least 2 requests, got 1

retry_behavior.retries_on_504

Expected at least 2 requests, got 1

retry_behavior.max_retries_respected

Expected 4 requests, got 1

deduplication.preserves_uuid_on_retry

Need at least 2 requests to check retry

deduplication.preserves_uuid_and_timestamp_on_retry

Expected at least 3 requests, got 1

deduplication.preserves_uuid_and_timestamp_on_batch_retry

Expected at least 2 requests, got 1

error_handling.retries_on_408

Expected at least 2 requests, got 1

Feature_Flags Tests

✅ 17/17 tests passed

View Details

Test	Status	Duration
Request Payload.Request With Person Properties Device Id	✅	518ms
Request Payload.Flags Request Uses V2 Query Param	✅	516ms
Request Payload.Flags Request Hits Flags Path Not Decide	✅	517ms
Request Payload.Flags Request Omits Authorization Header	✅	517ms
Request Payload.Token In Flags Body Matches Init	✅	516ms
Request Payload.Groups Round Trip	✅	517ms
Request Payload.Groups Default To Empty Object	✅	516ms
Request Payload.Disable Geoip False Propagates As Geoip Disable False	✅	516ms
Request Payload.Disable Geoip Omitted Defaults To False	✅	517ms
Request Payload.Flag Keys To Evaluate Contains Only Requested Key	✅	515ms
Request Lifecycle.No Flags Request On Init Alone	✅	513ms
Request Lifecycle.No Flags Request On Normal Capture	✅	512ms
Request Lifecycle.Two Flag Calls Produce Two Remote Requests	✅	518ms
Request Lifecycle.Mock Response Value Is Returned To Caller	✅	517ms
Retry Behavior.Retries Flags On 502	✅	619ms
Retry Behavior.Retries Flags On 504	✅	618ms
Side Effect Events.Get Feature Flag Captures Feature Flag Called Event	✅	523ms

[github-actions]

posthog-php-lib_curl Compliance Report

Date: 2026-10-23T13:00:21.621508+00:00
Duration: 116659ms

⚠️ Some Tests Failed

46/47 tests passed, 1 failed

---

Capture Tests

⚠️ 29/30 tests passed, 1 failed

View Details

Test	Status	Duration
Format Validation.Event Has Required Fields	✅	28ms
Format Validation.Event Has Uuid	✅	523ms
Format Validation.Event Has Lib Properties	✅	524ms
Format Validation.Distinct Id Is String	✅	526ms
Format Validation.Token Is Present	✅	525ms
Format Validation.Custom Properties Preserved	✅	525ms
Format Validation.Event Has Timestamp	✅	526ms
Format Validation.Non Utc Event Timestamp Is Converted To Utc	✅	525ms
Retry Behavior.Retries On 503	✅	5835ms
Retry Behavior.Does Not Retry On 400	✅	2527ms
Retry Behavior.Does Not Retry On 401	✅	2529ms
Retry Behavior.Respects Retry After Header	❌	6335ms
Retry Behavior.Implements Backoff	✅	16248ms
Retry Behavior.Retries On 500	✅	5634ms
Retry Behavior.Retries On 502	✅	5635ms
Retry Behavior.Retries On 504	✅	5633ms
Retry Behavior.Max Retries Respected	✅	17050ms
Deduplication.Generates Unique Uuids	✅	232ms
Deduplication.Preserves Uuid On Retry	✅	5631ms
Deduplication.Preserves Uuid And Timestamp On Retry	✅	10842ms
Deduplication.Preserves Uuid And Timestamp On Batch Retry	✅	5634ms
Deduplication.No Duplicate Events In Batch	✅	533ms
Deduplication.Different Events Have Different Uuids	✅	527ms
Compression.Sends Gzip When Enabled	✅	525ms
Batch Format.Uses Proper Batch Structure	✅	524ms
Batch Format.Flush With No Events Sends Nothing	✅	521ms
Batch Format.Multiple Events Batched Together	✅	513ms
Error Handling.Does Not Retry On 403	✅	2528ms
Error Handling.Does Not Retry On 413	✅	2526ms
Error Handling.Retries On 408	✅	5633ms

Failures

retry_behavior.respects_retry_after_header

Retry delay too short: 804ms < 2500ms

Feature_Flags Tests

✅ 17/17 tests passed

View Details

Test	Status	Duration
Request Payload.Request With Person Properties Device Id	✅	525ms
Request Payload.Flags Request Uses V2 Query Param	✅	522ms
Request Payload.Flags Request Hits Flags Path Not Decide	✅	523ms
Request Payload.Flags Request Omits Authorization Header	✅	523ms
Request Payload.Token In Flags Body Matches Init	✅	523ms
Request Payload.Groups Round Trip	✅	522ms
Request Payload.Groups Default To Empty Object	✅	523ms
Request Payload.Disable Geoip False Propagates As Geoip Disable False	✅	522ms
Request Payload.Disable Geoip Omitted Defaults To False	✅	523ms
Request Payload.Flag Keys To Evaluate Contains Only Requested Key	✅	523ms
Request Lifecycle.No Flags Request On Init Alone	✅	517ms
Request Lifecycle.No Flags Request On Normal Capture	✅	510ms
Request Lifecycle.Two Flag Calls Produce Two Remote Requests	✅	527ms
Request Lifecycle.Mock Response Value Is Returned To Caller	✅	524ms
Retry Behavior.Retries Flags On 502	✅	626ms
Retry Behavior.Retries Flags On 504	✅	627ms
Side Effect Events.Get Feature Flag Captures Feature Flag Called Event	✅	525ms

[veria-ai]

PR overview

All previously flagged issues have been addressed. No open security concerns remain on this pull request.

Security review

No open security issues remain on this pull request.

Fixed/addressed: 3 · PR risk: 0/10

[posthog]

PostHog Review · Standard · reviewed 2155f5a: 2 Should fix.

[posthog]

PostHog Review · Standard · reviewed e414589: 1 Must fix, 1 Should fix.

This is review round 3 on this pull request. When findings keep coming back in the same code, the approach is usually the problem. Before the next fix, check whether a different design would remove the whole class of findings.

[posthog]

PostHog Review · Standard · reviewed 7e232ad: 1 Must fix.

This is review round 4 on this pull request. When findings keep coming back in the same code, the approach is usually the problem. Before the next fix, check whether a different design would remove the whole class of findings.

[posthog]

PostHog Review · Standard · reviewed c5d55ce: 2 Should fix.

This is review round 5 on this pull request. When findings keep coming back in the same code, the approach is usually the problem. Before the next fix, check whether a different design would remove the whole class of findings.

This Standard review started from the posthog-review label. Comment @posthog review --deep for a Deep review, or @posthog help to see all commands.

[posthog]

PostHog Review · Standard · reviewed 5a4a540: 1 Must fix, 2 Should fix.

This is review round 6 on this pull request. When findings keep coming back in the same code, the approach is usually the problem. Before the next fix, check whether a different design would remove the whole class of findings.

[posthog]

PostHog Review · Standard · reviewed 577a8b7: 1 Must fix, 3 Should fix.

This is review round 7 on this pull request. When findings keep coming back in the same code, the approach is usually the problem. Before the next fix, check whether a different design would remove the whole class of findings.

[posthog]

PostHog Review · Standard · reviewed d16a8b1: 1 Must fix.

This is review round 8 on this pull request. When findings keep coming back in the same code, the approach is usually the problem. Before the next fix, check whether a different design would remove the whole class of findings.